Pinch to zoom, two fingers to move, on the timeline and the picture - #53
Merged
Conversation
There was no wheel handling anywhere. Zoom was three buttons and a tool, and a trackpad did nothing at all. A pinch reaches a page as a wheel event with `ctrlKey` set. There is no pinch event on a desktop and that is the only way a browser reports one, so that is what both surfaces listen for. On the timeline it zooms the time axis around the pointer, which is the whole reason to do it with a gesture: the buttons zoom around the left edge because they have no pointer to zoom around, and pinching around the left edge would send whatever you were looking at off the side of the screen. A plain two-finger scroll is left alone, because the viewport already scrolls both ways on its own and handling it here would move it twice. On the picture it zooms around the pointer through the same arithmetic the zoom tool already uses, and two fingers without ctrl move it -- but only once there is somewhere to move it, so at Fit the scroll is left to whatever is underneath rather than swallowed. Neither asks which tool is held. A gesture is not a tool: the hand and the zoom exist for a mouse, which has no way to say "closer" without one, and somebody pinching has already said it. The step is exponential rather than a multiplier, so the gesture is symmetric -- out and back by the same amount lands where it started. A multiplier does not: a tenth up and a tenth down is 0.99 of where it began. Six browser tests, sending the small repeated deltas a trackpad actually produces rather than one large jump. Two of the three faults were caught at once; the third took two goes and both misses are worth naming. Counting ruler labels could not see the multiplier drift, because the ruler picks a sensible interval and is therefore quantised: two per cent off looks identical. Measured on the drawn width instead. Then it still could not see it, because a clip opens fitted -- as far out as the timeline goes -- so the gesture out is clamped to the floor and lands on exactly where it started whether the arithmetic is symmetric or not. The test starts part way in now, and the fault fails it: 1638.5px became 1606.9px. 359 unit tests, 96 browser tests, build clean. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ux1kydvUkLRoMbHp82ofDi
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
There was no wheel handling anywhere in the app. Zoom was three buttons and a tool; a trackpad did nothing at all.
A pinch reaches a page as a wheel event with
ctrlKeyset — there's no pinch event on a desktop, and that's the only way a browser reports one. Both surfaces listen for it.What it does
Timeline — pinch zooms the time axis around the pointer. That's the whole reason to do it with a gesture: the buttons zoom around the left edge because they have no pointer to zoom around, and pinching around the left edge would send whatever you were looking at off the side of the screen. Plain two-finger scroll is left alone — the viewport already scrolls both ways on its own, and handling it here would move it twice.
Picture — pinch zooms around the pointer through the same arithmetic the zoom tool already uses; two fingers without ctrl move it, but only once there's somewhere to move it. At Fit the scroll passes through rather than being swallowed.
Neither asks which tool is held. A gesture is not a tool: the hand and the zoom exist for a mouse, which has no way to say "closer" without one. Somebody pinching has already said it.
The step is exponential rather than a multiplier, so the gesture is symmetric — out and back by the same amount lands where it started.
Six tests, and two misses worth naming
They send the small repeated deltas a trackpad actually produces, not one large jump. Two of three injected faults were caught immediately. The third — replacing the exponential with a straight multiplier — took two attempts:
Starting the test part-way in fixes it, and the fault now fails plainly:
359 unit tests, 96 browser tests (was 90), build clean.
Still to come from your request: GIF/APNG/animated-WebP import via
ImageDecoder, and the "without lagging" work for multi-minute 1080p.🤖 Generated with Claude Code
https://claude.ai/code/session_01Ux1kydvUkLRoMbHp82ofDi
Generated by Claude Code